Background: The occurrence of epithelial-mesenchymal transition (EMT) and immune evasion is considered to contribute to poor prognosis in lung adenocarcinoma (LUAD). Therefore, this study aims to explore the key oncogenes that promote EMT and immune evasion and reveal the expression patterns, prognostic value, and potential biological functions. Methods: Firstly, we identified gene modules associated with EMT and Tumor Immune Dysfunction and Exclusion (TIDE) through weighted gene co-expression network analysis (WGCNA). Next, we utilized differential analysis and machine learning to identify the key genes and validate them. Moreover, we analyzed the correlation between key genes and tumor microenvironment remodeling, drug sensitivity, as well as mutation frequency. Furthermore, we explored and validated their malignant biological characteristics through in vitro experiments and clinical samples. Finally, potential drugs for LUAD were screened based on CMap and validated through experiments. Results: Firstly, WGCNA analysis revealed that red and green modules were highly correlated with EMT and TIDE. Among them, upregulated expression of SPOCK1 was observed in lung adenocarcinoma tissues and was associated with poor prognosis. Additionally, patients in the high SPOCK1 group showed more activation of malignant oncogenic pathways, higher infiltration of immunosuppressive components, and a higher frequency of mutations. The knockdown of SPOCK1 suppressed invasion and metastasis capabilities of lung adenocarcinoma cells, and the high expression of SPOCK1 was associated with low infiltration of CD8+ T cells. Therapeutic aspects, SPOCK1 can be a candidate indicator for drug sensitivity and CMap showed that VER-155008 was the drug candidate with the largest perturbation effect on the SPOCK1 expression profile. In vitro and in vivo experiments validated the cancer-inhibitory effect of VER-155008 in LUAD. Conclusion: This study revealed through comprehensive bioinformatics analysis and experimental analysis that SPOCK1 can promote EMT and immune escape in LUAD, and it may serve as a promising candidate prognostic biomarker and therapeutic target for LUAD.
Objection: Investigating the key genes and mechanisms that influence stemness in lung adenocarcinoma. Methods: First, consistent clustering analysis was performed on lung adenocarcinoma patients using stemness scoring to classify them. Subsequently, WGCNA was utilized to identify key modules and hub genes. Then, machine learning methods were employed to screen and identify the key genes within these modules. Lastly, functional analysis of the key genes was conducted through cell scratch assays, colony formation assays, transwell migration assays, flow cytometry cell cycle analysis, and xenograft tumor models. Results: First, two groups of patients with different stemness scores were obtained, where the high stemness score group exhibited poor prognosis and immunotherapy efficacy. Next, LASSO regression analysis and random forest regression were employed to identify genes (PBK, RACGAP1) associated with high stemness scores. RACGAP1 was significantly upregulated in the high stemness score group of lung adenocarcinoma and closely correlated with clinical pathological features, poor overall survival (OS), recurrence-free survival (RFS), and unfavorable prognosis in lung adenocarcinoma patients. Knockdown of RACGAP1 suppressed the migration, proliferation, and tumor growth of cancer cells. Conclusion: RACGAP1 not only indicates poor prognosis and limited immunotherapy benefits but also serves as a potential targeted biomarker influencing tumor stemness.